PDF Compression: When Quality Matters vs When Size Matters

Here's a conversation that happens way too often: Someone compresses a PDF down to 2MB to email it, then complains that the charts are blurry and the text looks fuzzy. On the flip side, someone else keeps a 150MB PDF of scanned documents because they're terrified of losing quality, then can't upload it anywhere.

The problem isn't compression itself. It's that people treat it like a binary choice: either keep the full file, or shrink it and accept whatever damage happens. But PDF compression isn't a one-size-fits-all problem. It's a trade-off, and different use cases justify different trade-offs.

The Real Cost of Compression

Before we get into the framework, let's talk about what compression actually does. There are basically two types of PDF compression happening:

• Lossless compression — removes redundancy and unused data without discarding information. Metadata cleanup, font subsetting, removing hidden layers. File gets smaller but quality doesn't change.
• Lossy compression — actually throws data away. Reducing image resolution, converting to JPEG, lowering color depth. Smaller file, but you can't get that data back.

Most compression tools do both. They clean up metadata first (good), then resample images down to lower DPI (bad for quality). The more aggressive you go, the more lossy compression kicks in.

Here's the thing: once lossy compression happens, it's permanent. You can't un-JPEG a PDF. You can't increase resolution that's been thrown away. This is why the decision matters.

The Decision Framework

Instead of asking "how small can I make this?", ask these questions in order:

Question 1: Do I need the original?

This is the gatekeeper question. If you're compressing a working copy, not a master, aggressive compression is way safer. You can always go back to the original.

But if this is your only copy, or if you might need to re-edit it later, keep the original uncompressed. Full stop. Make a compressed copy for distribution, but don't overwrite your master.

Question 2: What's the bottleneck?

Size matters for specific reasons. Is it:

• Email limits? (usually 20-25MB max) — Moderate compression is fine. Email rarely needs pristine quality.
• Upload to cloud storage? (usually no real limit) — Question whether you need compression at all.
• Mobile bandwidth? (slow connections, large files frustrate users) — Compression is worth it here.
• Printing or scanning? (you need quality) — Avoid aggressive compression.
• Archival storage? (you're keeping this forever) — Usually not worth compressing.

Different bottlenecks have different solutions. Email doesn't need 300 DPI images, but a design portfolio absolutely does.

Question 3: What's in the PDF?

Content type matters hugely:

• Mostly text? Lossless compression gives you huge savings (50-70% smaller) with zero quality loss. Text is weird to compress and benefits massively from cleanup.
• Lots of photos? You're limited. Photos don't compress well. Resampling to 150 DPI might only save 20-30%. Going lower gets blurry fast.
• Scanned documents? These are brutal. Scans are basically uncompressable. You're either keeping them full-quality (huge) or converting to JPEG (ugly).
• Mixed content? Different parts need different treatment. Charts stay crisp at 200 DPI. Photos can go lower. Text never needs compression.

This is why one compression setting doesn't fit everything. A text PDF and a photo PDF are completely different problems.

Compression Strategies by Use Case

Let's get specific. Here's what to actually do:

Email Distribution (non-sensitive)

Target: Under 10MB. Strategy: Aggressive lossless + moderate lossy.

• Remove metadata, embedded fonts, hidden layers (lossless)
• Resample images to 150 DPI
• Convert photos to JPEG compression
• Test on sample pages first

You'll typically hit 70-80% size reduction. Quality looks fine on screen, maybe not for printing. Perfect for "here's your invoice" scenarios.

Legal/Business Documents (archival)

Target: Keep original, distribute uncompressed. Strategy: Zero loss.

• Don't compress. Period.
• Store on proper file systems with redundancy
• If space is truly an issue, make a separate compressed version and label it clearly

The cost of keeping 50MB is pennies. The cost of a blurry contract signature is catastrophic. This is not where you optimize.

Web Distribution (portfolios, lookbooks)

Target: 3-5MB. Strategy: Careful image optimization.

• Start with lossless compression
• Resample images to 200 DPI (web standard)
• Keep JPEG compression moderate (80-85%)
• Test on actual website to see how people will view it

This usually gives 50-60% reduction while keeping quality acceptable for screen viewing. Images still look professional.

Scanned Documents (practical use)

Target: Depends on whether you'll re-scan. Strategy: Depends on use.

• If it's just reference? Compress to 50-100MB, acceptable quality is fine
• If you'll need clarity later? Don't compress. Re-scan later if needed
• If it's research material? Archive at full quality, make a compressed copy for mobile viewing

Scans don't benefit much from compression anyway. You usually get 20-30% reduction and lose 50% of perceived quality. Not a good trade-off.

Tools That Do It Right

Most compression tools are brutally aggressive by default. They smash your file down and don't tell you what they sacrificed. Better tools let you control the trade-off:

• KokoConvert — Lets you choose compression level (mild, moderate, aggressive) before you convert. You see the size estimate first.
• Built-in OS tools — macOS Preview and Windows "Reduce File Size" are fine for casual use, but zero control.
• Ghostscript — Powerful but requires command line. Full control, but technical.

Use KokoConvert's PDF compression tool if you want a balance of simplicity and control. You can see exactly what compression level does to your file before committing.

The Practical Workflow

Here's what I actually do:

1. Keep the original uncompressed in archive storage
2. Ask "why do I need compression?" — usually the answer reveals what level is actually needed
3. Try lossless compression first, measure the result
4. If still too large, try 200 DPI resampling as middle ground
5. Test the result with actual viewers on actual devices (email, phone, print preview)
6. Only then decide if aggressive JPEG conversion is worth it

This takes maybe 5 extra minutes but saves you from sending a blurry document or discovering your "compressed" file is still 60MB because it was full of embedded fonts.

What About Size Estimates?

When you're about to compress something, can you predict the final size? Sort of. Here are rough estimates:

• Lossless only: 50-70% reduction (depends heavily on content)
• Lossless + 200 DPI resampling: 60-75% reduction
• Lossless + 150 DPI resampling: 70-80% reduction
• Aggressive JPEG: 80-90% reduction (quality drops significantly)

But honestly, these are guesses. The only way to know is to test on a sample. A 50MB text PDF might compress to 5MB. A 50MB photo PDF might only get to 30MB. Same starting size, wildly different results.

This is why good compression tools show you a preview or estimate first. You shouldn't have to guess and iterate.

One Last Thing: Document Your Decisions

If you compress something and keep both versions, label them clearly. "invoice-original.pdf" and "invoice-email.pdf" so future-you knows which is which. Or better yet, just keep one version and compress on-demand when needed.

The worst scenario is six months later wondering if you have the original or just a blurry compressed copy. Take 10 seconds to name files properly.

PDF compression isn't complicated. It's just trade-offs. Know your bottleneck, understand your content, and make decisions that fit your actual use case. That's it. You don't need tiny files for everything, and you shouldn't sacrifice quality where it matters.